4
$\begingroup$

Take two (non-zero) Hermitian operators $A$ and $B$. I want to proof that there exists no unitary operator $W$ such that: $$W^{\dagger}AW = A + B$$ For my research I proved this for some specific case that involves symmetries. However I would like to know whether there is a general proof of this statement. If not, are there any conditions on $A$ and $B$ so it is true.
So far I tried solving it with properties of the trace and unitary gates: \begin{align} Tr(A) = Tr(A) + Tr(B) &\rightarrow Tr(B) = 0 \\ Tr(A^2) = Tr(A^2) + Tr(B^2) + 2\cdot Tr(AB) &\rightarrow Tr(B^2) + 2\cdot Tr(AB) = 0 \end{align} And so on. But this does not give me any contradiction e.g. $B = 0$ or something
Hence, my question is can $A$ and $A+B$ be unitarily equivalent?

Edit: included the non-zero condition

$\endgroup$
3
  • $\begingroup$ I don't think that this is true in general. Take any hermitian $A$ and $B=0$ as well as $W=\mathbb I$, for example. $\endgroup$ Commented Jan 10, 2023 at 11:49
  • $\begingroup$ It is also true if $A \neq 0$ and $B = 0$. However this is trivial. $\endgroup$
    – stacksper
    Commented Jan 10, 2023 at 12:02
  • 1
    $\begingroup$ If you restrict yourself to finite-dimensional matrices the answer is given by en.wikipedia.org/wiki/Specht%27s_theorem. $\endgroup$
    – Martin C.
    Commented Jan 10, 2023 at 14:22

3 Answers 3

7
$\begingroup$

The time evolution of the position operator of a free particle in quantum mechanics (Hamiltonian $H=P^2/2m$) is a nice counterexample:

$ e^{iHt/\hbar} X e^{-iHt/\hbar} = X + P t/m$,

with $A=X$ (position operator at time $t=0$), $B= P t/m$ (momentum operator $P$ times $t/\hbar$) and $W= \exp(-iHt/\hbar)$ (time evolution operator).

$\endgroup$
6
$\begingroup$

Here is a simple counter-example: $$e^{iaP/\hbar}Xe^{-iaP/\hbar}=X+a$$ where $X$ is the position operator, $P$ is the momentum operator, and $a$ is a constant.

$\endgroup$
4
$\begingroup$

A nice counterexample has already been presented.

More generally, for any $W$ of the form $e^{X}$, where $X$ is an element of a Lie algebra, it follows from an identity related to the BCH formula that

$e^X A e^{-X} = \sum_{n=0}^{\infty}\frac{[(X)^n, A]}{n!}$.

So in such cases (as in the examples) the statement will clearly fail if $A$ and $X$ do not commute, and their commutator is also Hermitian.

$\endgroup$
2
  • $\begingroup$ This doesn't affect the main point, but I find the notation $[X^n,A]$ for $ad_X^n(A)$ rather misleading. (In the Wikipedia article they write $[(X)^n,A]$ which is maybe somewhat better...) $\endgroup$
    – Noiralef
    Commented Jan 11, 2023 at 4:35
  • 1
    $\begingroup$ @Noiralef fair comment, I have edited the post $\endgroup$
    – Martin C.
    Commented Jan 11, 2023 at 6:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.